Summary We are interested in exploring the interactions between the mosquito (A. gambiae and A. stephensi) immune system and the malaria parasite and to understand how they determine vector competence. Some major areas of interest include: The role of reactive oxygen species (ROS) in determining the balance between fecundity, immunity and vectorial capacity in A. gambiae. The susceptibility to bacterial infection of three different A. gambiae strains that differ in their steady-state levels of H2O2 in hemolymph and in their susceptibility to malaria infection correlated directly with their systemic levels of H2O2. Furthermore, reduction of ROS by dietary supplementation of Vitamin C (strong antioxidant), dramatically decreased survival following bacterial challenge. Midgut mRNA levels of Mn superoxide dismutase (MnSOD), Cu/ZnSODs and glutathione peroxidase (GPx) increased in mosquitoes that received either a Plasmodium-infected or an uninfected blood meal, indicating that blood feeding induces oxidative stress. Surprisingly, midgut catalase mRNA expression was induced in mosquitoes fed uninfected blood but not in Plasmodium-infected mosquitoes. Midgut catalase activity, measured 24 hr after a blood meal was also lower in Plasmodium-infected mosquitoes. This decrease in catalase expression is triggered by ookinete invasion, as mosquitoes infected with a parasite strain in which ookinetes are unable to invade the epithelium induce catalase. Reduction of catalase expression by dsRNA silencing significantly decreased the number of developing oocysts, suggesting that the increased levels of H2O2 limit Plasmodium infection. Thioredoxin reductase mRNA induction in response to blood feeding is also suppressed in Plasmodium-infected midguts and potentiates the effect of reduced catalase expression in epithelial cells by preventing thioredoxin-mediated detoxification of H2O2 . A manuscript describing these findings is in preparation. Plasmodium-nvaded cells undergo a peroxidase-mediated nitration reaction, and increased availability of H2O2 may be necessary during this process, as this reaction uses nitrate and H2O2 as substrates (see below). The effect of aging on reproductive output in A. gambiae females from three strains that differ in their ability to melanize P. falciparum and in their systemic levels of H2O2 was analyzed. The number of developing oocytes following a blood meal decreases with age in all strains, but this decline was much more pronounced in G3 (unselected, intermediate H2O2 levels) and R (refractory to Plasmodium infection, high H2O2) than in the S (highly susceptible to Plasmodium, low H2O2) strain. Reduction of ROS levels in G3 and R females by administration of anitoxidants prevents this age-associated decline in fecundity. The S and G3 strains are fixed for two functionally different catalase alleles, Ser/Ser (high activity) and Trp/Trp (lower activity), respectively. Biochemical analysis of recombinant proteins revealed that the Trp is a less efficient enzyme with lower Vmax and higher Km. The replacement of Ser for Trp appears to destabilize the active tetrameric form of the enzyme. Fecundity studies in the R strain, in which both catalase alleles are present revealed a strong association between reproductive output and the catalase phenotype of individual females. Trp/Trp females produced significantly less viable larvae than Ser/Ser homozygotes and Trp/Ser females had an intermediate level. A systemic reduction in catalase activity by RNAi knockdown significantly reduced the reproductive output, indicating that catalase plays a central role in protecting the oocyte and early embryo from ROS damage. A manuscript describing these results has been submitted to PNAS. Cell biology and biochemistry of the defense responses of midgut epithelial cells to ookinete invasion: Parasite invasion in S females triggers a series of toxic reactions, such as the generation of nitrogen dioxide, that result in apoptosis; and ookinetes have a limited time-window to escape unharmed from the invaded cell. An inducible peroxidase mediates nitration invaded cells. Apoptotic responses and tissue repair are accelerated in the R strain, probably due to the higher steady-state levels of H2O2 . When the expression of heme peroxidases was compared between S and R females, a heme peroxidase that is down regulated by Plasmodium infection in G3 remained expressed at higher levels in R females, while a second peroxidase that is induced by Plasmodium infection in the S strain was found to be constitutively induced in the R strain. Silencing of either one of these genes by systemic dsRNA injection reverted the refractory (R) phenotype, so that a significantly higher percent of healthy developing oocysts were observed 48 h post-infection. We have named these two proteins ?encapsulation-mediating peroxidase 1? (Emper1) and Emper2, respectively. Emper1 is highly induced by blood feeding and is secreted into the midgut lumen. Surprisingly, Emper 1 silencing in S females is detrimental to the parasite and dramatically reduces Plasmodium infectivity. This enzyme appears to have some general protective effect in the midgut lumen, as silencing also reduced bacterial proliferation. Experiments with cultured ookinetes exposed to different treatments in vitro confirmed that, under conditions of low H2O2, peroxidases have a protective effect, while they are deleterious to the parasite when H2O2 levels are high. The role of STAT signal transduction pathways on the regulation genes mediating mosquito immune responses to Plasmodium. We have identified two transcription factors from the signal transduction and activation of transcription (STAT) family in A. gambiae, AgSTAT-A and AgSTAT-B; as well as orthologues of two suppressors of these pathways in vertebrates, SOCS and PIAS. All these genes are expressed in the immune-responsive A. gambiae Sua 5.1 cell line and NOS and SOCS are transcriptionally activated in response to bacterial challenge. These genes are also induced in Plasmodium-infected mosquitoes. Silencing of either AgSTAT-A or AgSTAT-B by transfecting cells with in vitro synthesized dsRNA resulted in a significant decrease of NOS and SOCS mRNA expression following bacterial challenge, indicating that both transcription factors participate in the regulation of NOS and SOCS expression. In vivo silencing of AgSTAT-B by dsRNA injection decreased NOS and SOCS mRNA expression in response to Plasmodium, demonstrating the central role of this transcription factor in the regulation of both genes. Preliminary data indicate that silencing AgSTAT-A or AgSTAT-B in S females has opposite effects. Parasites numbers increased following AgSTAT-A silencing, while they decreased when AgSTAT-B expression was reduced. Silencing of SOCS dramatically reduced the number of developing oocytes. The differences in expression following AgSTAT-A and AgSTAT-B silencing will be evaluated using microarray analysis. These data indicate that the STAT pathways regulate defense responses that limit parasite development. Identification of mosquito genes required for Plasmodium oocyst development: Dr. David Schneider from Stanford University performed a genetic screen to identify genes from Drosophila that were necessary to allow development of injected Plasmodium gallinaceum ookinetes into mature oocysts. He identified 20 mutations that affected parasite development. In a collaborative project, we have cloned the A. gambiae orthologues of five of these genes and have silenced them by dsRNA injection. Four of the five genes tested also affected Plasmodium berghei development after natural infections (feeding on infected mice) in A. gambiae. The effect of silencing these genes on P. falciparum development is currently under evaluation.